Add Checkpointing to profiling to save profiling progress

[IanBogle]

Is your feature request related to a problem? Please describe.
Heavy weight benchmarks that take a significant time to profile may be cut short by time limits on scheduled systems, resulting in losing all profiling data

Describe the solution you'd like
Add a feature that checkpoints profiling progress when an omniperf run gets interrupted, which will save time when resuming a profile.

Describe alternatives you've considered
Running several filtered profile runs on IP blocks could emulate this feature, but has significant manual overhead for users. Ideally checkpointing would happen transparently